Location: Alcott Boardroom
This is a roundtable in the Cyberlearning 2017 Roundtable session.
Providing feedback on complex tasks: Algorithmic formative assessment to help students learn the experimental process
Under a recently completed NSF Cyberlearning grant we investigated ways of providing students feedback on their understanding of complex, open-ended tasks. The primary approach was to put light constraints on the tasks that made it easier to design algorithms for recognizing student confusions. One culmination of this project was a new virtual lab called Understanding Experimental Design which presents students with a problem (cute critters are dying!) and asks them to design and carry out experiments to identify a cause. I’ll show the open-ended experimental design piece of the lab and the types of confusions we’re able to capture and provide students feedback on. I’ll also show some results from studies we’ve done looking at the effect of the constraints and the feedback on student learning. I’d welcome a discussion of how this approach might generalize to other higher-order skills that we ask students to learn in large classes where direct instructor feedback is often lacking.
Project: Homepage, NSF Award #1227245 – DIP: Using dynamic formative assessment models to enhance learning of the experimental process in biology
Inq-Blotter: A real time alerting tool to transform teachers’ assessment of science inquiry practices
Teachers are often frustrated because large-scale assessment data are neither timely nor adequate to understand students’ needs so they can improve science instruction/learning. To address this challenge, we propose to extend, pilot, implement, and study Inq-Blotter, a scalable, data-driven, web-based alerting system that can revolutionize teachers’ formative assessment practices and instruction of inquiry. Inq-Blotter is a dashboard that will let teachers know who needs the most help and on which inquiry practices, closing the formative assessment loop by providing real time skill data to teachers and affording rich scientific discourse between teachers and their students. Inq-Blotter will be integrated with Inq-ITS (Inquiry Intelligent Tutoring System) for middle school Physical Science, a learning environment aligned to the newly released Next Generation Science Standards in which students show what they know by conducting inquiry within virtual labs. Unique to Inq-ITS is its ability to automatically assess inquiry using patented algorithms based on knowledge-engineering and data mining, making alerting possible. In the proposed project, we will implement Inq-Blotter in our classroom partner schools and analyze the discourse between teachers and students for each NGSS practice to study how this new genre is used in real classroom-based formative assessment.
Project: Homepage, NSF Award #1629045 – Inq-Blotter: A real time alerting tool to transform teachers’ assessment of science inquiry practices
Automated Guidance for Student Written Explanations
I would like to share research on the design of automated guidance for student writing in online inquiry science projects. I will present the design of the explanation questions and NLP scoring rubrics, embedded into 1-2 week long projects. I will present the iterative refinement of the technology design with results from comparison studies conducted in middle school classrooms to identify promising ways to guide students, and to alert the teachers in real-time to which students are in need of help.
Project: Homepage, NSF Award #1451604 – Project Learning with Automated, Networked Support (PLANS)